ATMOSPHERIC ELECTRICITY. 
143 
the Orinoco, in the 'Rio Atabapo- Did they penetrate 
*nto the centre of equinoctial America from the mouth of 
the Amazon, by the communication of that river with the 
Itio Negro, the Oassiquiare, and the Orinoco ? They are 
found here at all seasons, and nothing seems to denote 
that they make periodical migrations like salmon. 
While the thunder rolled around us, the sky displayed 
or dy scattered clouds, that advanced slowly toward the 
zenith, and in an opposite direction. The hygrometer of 
Deluc was at 53°, the centigrade thermometer 23’7°, and 
' fissure’s hygrometer S7‘5°. The electrometer gave no 
w gn of electricity. As the storm gathered, the blue of the 
sk -y changed at first to deep azure and then to grey. The 
'esicular vapour became visible, and the thermometer rose 
three degrees, as is almost always the case, within the 
tropics, from a cloudy sky which reflects the radiant heat of 
he soil, y heavy rain fell. Being sufficiently habituated 
0 the climate not to fear the effect of tropical rains, we 
Remained on the shore to observe the electrometer. I held 
u more than twenty minutes in my hand, six feet above the 
8 r ound, and observed that in general the pith-balls separated 
011 'y a few seconds before the lightning was seen. The 
Reparation was four lines. The electric charge remained 
he same during several minutes ; and having time to deter- 
mine the nature of the electricity, fey approaching a stick of 
8 ??nng-wax, I saw here what I had often observed on the 
H o e of the Andes during a storm, that the electricity of 
atmosphere was first positive, then nil, and then ne- 
gative. These oscillations from positive to negative were 
m ® n re peated. Yet the electrometer constantly denoted, a 
j t e before the lightning, only E., or + E., and never — E. 
°-ards the end of the storm the west wind blew very 
to r b'be clouds dispersed, and the thermometer sunk 
fro ^ on account of the evaporation from the sod, and the 
radiation towards the sky. 
of ave en t er ed into these details on the electric charge 
tile . ^mospbere because travellers in general confine 
pelves to the description of the impressions produced 
tro a T r0peal1 newl y arrived by the solemn spectacle of a 
P’cal storm. In a country where the year is divided into 
